Thompson Sampling in Partially Observable Contextual Bandits.
Hongju ParkMohamad Kazem Shirani FaradonbehPublished in: CoRR (2024)
Keyphrases
- partially observable
- reinforcement learning
- markov decision processes
- state space
- decision problems
- dynamical systems
- multi armed bandit
- partial observability
- infinite horizon
- markov decision problems
- action models
- partial observations
- belief state
- partially observable domains
- reward function
- partially observable environments
- long run
- machine learning
- policy iteration
- belief space
- planning domains
- dynamic systems
- heuristic search
- np hard